Intel’s foundry turnaround runs on a big promise, but outside revenue remains small

Intel’s $500 billion valuation jump is forcing a blunt question: where are the “proof points” that Intel Foundry can become a major contract manufacturer, not just a production line for Intel’s own products. In a July 7 market analysis, Trefis said Intel’s stock has risen more than fivefold over the past 12 months, moving from roughly the high-teens to a recent range near $120, and added nearly $500 billion to its market capitalization.

The optimism, according to the analysis, rests on two intertwined narratives. One is stronger CPU demand tied to AI workloads, with server processors increasingly seen as central to the AI buildout. The other, and more specific to the foundry business, is a bet that Intel’s manufacturing arm can scale output and attract enough outside customers to change the economics of the company’s semiconductor production.

Intel Foundry, as framed by the market commentary, is supposed to evolve from what investors sometimes describe as an internal cost center into a world-class external business. That shift matters because contract manufacturing would diversify demand beyond Intel’s own chip portfolio and potentially create more stable, foundry-led revenue streams. The analysis also points to a broader industry theme, namely that customers looking for alternatives to Asia-based leading-edge manufacturing capacity could view an Intel foundry model as a route to geographic diversification.

But the reported numbers highlight the gap between the narrative and the scale of external traction. For Q1 2026, Trefis cited total foundry revenue of $5.4 billion, up from $4.7 billion a year earlier. External foundry revenue was just $174 million, while the foundry posted an operating loss of $2.4 billion, roughly flat versus a $2.3 billion loss a year earlier.

For the full year 2025, the same analysis said total foundry revenue was $17.8 billion, with external contribution of only $307 million. It also cited an operating loss of $10.3 billion for the year. In other words, even with growing overall foundry sales, the economics still appear to be dominated by internal demand, not by a meaningful base of outside customers paying for advanced process capacity.

The company’s challenge, as reflected in the commentary, is that manufacturing chips for Intel validates the technology but does not automatically deliver foundry economics. External customers are the critical variable for a contract-manufacturing story, because they determine whether leading-edge capacity can be monetized as a business in its own right rather than funded largely by Intel’s own strategy.

Sector context adds to the tension. The market is reacting not only to current chip demand but to expectations about the future manufacturing landscape for advanced semiconductors. If Intel can credibly convert its manufacturing roadmap into repeatable volume for external partners, the foundry could become a durable revenue engine. If not, investors may eventually reassess whether the company can close the gap between foundry optimism and the pace of outside commercialization.

Still, there is a key limitation in what is publicly summarized in the market post: it does not provide granular detail on the specific customers, product nodes, or contract milestones that would constitute the clearest “proof points.” The analysis also does not address whether losses are being driven primarily by ramp costs, yield learning, capital intensity, or mix of internal versus external workloads. Without that detail from Intel’s filings or official updates, it remains difficult to separate near-term accounting losses from the underlying progress of the manufacturing operation.